How to Use the Webflow MCP in Pydantic AI

Q: How do I ensure data integrity when listing Webflow collections with Pydantic AI?

When using listcollectionitems and listcollections, Pydantic AI validates the structure of every returned item. If a field is missing or typed wrong, your agent catches it immediately, preventing faulty code execution.

Q: What happens if I use the Pydantic AI framework to publish a site?

The system checks all required parameters for publishsite against defined models. If something is malformed—say, an invalid domain name—the agent fails with a clear validation error instead of submitting bad data.

Q: Does the Pydantic AI framework help audit Webflow sites?

Yes. By using getsite or listpages, your agent gets typed, validated objects for site details and page lists. This allows you to write reliable, deterministic code that processes web assets.

Q: What specific Webflow domains can the Pydantic AI validate?

The listdomains tool returns domain strings which are then processed through a type-safe model. This guarantees they look like valid domain names before your agent uses them.

Q: What types of data does this MCP Server touch with Pydantic AI?

It handles site details, user information (getuser), and structured content (CMS items). The core benefit is that all these varied data types are validated against strict runtime models.

Guarantee Webflow actions are correct using the Pydantic AI framework.

See Vinkius in Action

Works with every AI agent you already use

…and any MCP-compatible client

MCP Servers - Free for Subscribers

Connect Webflow MCP to Pydantic AI

Create your Vinkius account to connect Webflow to Pydantic AI and route execution through our secure gateway. The platform manages server hosting, runtime updates, and security layers. Configuration requires no manual server provisioning.

GDPR Free for Subscribers

Setup Webflow with Pydantic AI

Ask AI about this MCP

ChatGPT

Claude

Perplexity

Reliable CMS Writes with Pydantic AI

When you call `create_collection_item`, your agent's output is immediately validated against a schema. This means the Webflow data written—like text fields or dates—won't corrupt the collection because of bad input. The framework enforces correctness. This prevents silent failure; if the API returns unexpected data, Pydantic AI fails loudly, alerting you that the structured content wasn't right.

Type-Safe Site Publishing via MCP Server

Running `publish_site` requires specific parameters. With this setup, your agent ensures every required field (like the target environment or version ID) is correctly provided before executing the command. You don't risk a partially published site. This level of runtime validation makes deployment reliable—the output must conform to the expected Webflow structure.

Managing Site Metadata with Pydantic AI

Tools like `get_site` and `list_pages` return data that is immediately parsed into Python objects. This means your agent doesn't just get a JSON blob; it gets guaranteed fields and types for site details, making downstream processing much safer. It provides confidence that the metadata—be it page names or domain records—is exactly what you expect.

Setup guide

Set up Webflow MCP in Pydantic AI

Prerequisites

Python 3.10+ installed
pydantic-ai-slim[fastmcp] package
Active Vinkius subscription with a valid endpoint token

1

Install Pydantic AI with FastMCP
Run pip install "pydantic-ai-slim[fastmcp]". The FastMCP toolset replaces the deprecated MCPServerHTTP class with full protocol support.
2

Configure the FastMCPToolset
Pass a JSON-style config dict to FastMCPToolset with your Vinkius URL. Replace [YOUR_TOKEN_HERE] with your token from cloud.vinkius.com. Supports Streamable HTTP, SSE, and Stdio transports.
3

Create and run your agent
Pass the toolset to Agent(toolsets=[toolset]) and call agent.run(). Swap openai:gpt-4o for any supported model — Anthropic, Google, Mistral, or Groq.

agent.py

from pydantic_ai import Agent
from pydantic_ai.toolsets.fastmcp import FastMCPToolset

toolset = FastMCPToolset({
    "mcpServers": {
        "webflow-mcp": {
            "url": "https://edge.vinkius.com/[YOUR_TOKEN_HERE]/mcp"
        }
    }
})

agent = Agent(
    "openai:gpt-4o",
    toolsets=[toolset],
    system_prompt="You have access to Webflow tools.",
)

result = await agent.run("List recent Webflow transactions")
print(result.output)

Get your connection token →

Independent Platform Disclaimer: Vinkius is an independent platform and is not affiliated with, endorsed by, sponsored by, verified by, or otherwise authorized by Webflow. All third-party trademarks, logos, and brand names are the property of their respective owners. Their use on this website is strictly for informational purposes to identify service compatibility and interoperability.

Why Choose Vinkius

Vinkius connects your tools to AI with real-time monitoring and automatic cost savings — all from one dashboard.

Connect Webflow now

Real-time monitoring

Live

visibility into every interaction

Connect your favorite tools to your AI and see exactly what's happening — every request, every response, in real time.

Built-in savings

60%

lower AI costs

Vinkius compresses data between your apps and your AI automatically. Lower bills every month — no configuration required.

Single dashboard

One

place for every integration

Every tool your AI connects to, managed from a single screen. One account, complete control.

Common questions about Webflow MCP in Pydantic AI

When using `list_collection_items` and `list_collections`, Pydantic AI validates the structure of every returned item. If a field is missing or typed wrong, your agent catches it immediately, preventing faulty code execution.

The system checks all required parameters for `publish_site` against defined models. If something is malformed—say, an invalid domain name—the agent fails with a clear validation error instead of submitting bad data.

Yes. By using `get_site` or `list_pages`, your agent gets typed, validated objects for site details and page lists. This allows you to write reliable, deterministic code that processes web assets.

The `list_domains` tool returns domain strings which are then processed through a type-safe model. This guarantees they look like valid domain names before your agent uses them.

It handles site details, user information (`get_user`), and structured content (CMS items). The core benefit is that all these varied data types are validated against strict runtime models.

Use it with your favorite AI tools

Connect this server to Cursor, Claude, VS Code, and more.

OpenAI Agents SDK sdk-python

Google ADK sdk-python

Pydantic AI sdk-python

Vercel AI SDK sdk-typescript